The present invention relates to an autochanger of writable discs.
FIG. 4 shows a conventional autochanger of writable discs. In this figure, a writable disc of, for example, a write-once type optical disc 1 is a large capacity disc medium having storage capacity of several hundred mega-bytes to several giga-bytes per one disc and several discs to several hundred discs are accommodated in order in a disc accommodation unit 2. In the predetermined area of the optical disc 1, defect management information showing addresses of defective sectors in the optical disc 1 and addresses of replacement sectors for the defective sectors is recorded. A loading mechanism 3 automatically picks up the predetermined optical disc contained in the disc accommodation unit 2 and transfers the picked-up optical disc to apply it to a recording/reading device 4, automatically removes the optical disc from the recording/reading apparatus 4 and returns it to the disc accommodation unit 2. The recording/reading device 4 records information into the optical disc installed therein and reads the information recorded on the optical disc. A plurality of the recording/reading devices 4 may be used in the autochanger.
A controller 5 consists of a microcomputer and performs such operations according to commands supplied from a host computer 6 through an interface 7 (SCSI or the like) as; controlling the loading mechanism 3 so as to load the suitable optical disc from the disc accommodation unit 2 to the recording/reading device 4, controlling the recording/reading device 4 so as to record the information supplied from the host computer 6, and reversely controlling the recording/reading device 4 so that the recorded information is read out of the optical disc and transferred to the host computer 6. The controller 5 contains a RAM (random access memory) for storing defect management information of the optical disc.
When a new optical disc has been installed in the recording/reading device 4 in response to the commands sent from the host computer 6, the controller 5 instructs the recording/reading device 4 so as to read the defect management information recorded in the predetermined area of the optical disc and writes the read-out defect management information into the RAM 8 in the controller 5. After that, when the controller 5 receives the command sent from the host computer 6 ordering to read the information recorded in the optical disc, the controller 5 judges from the defect management information in the RAM 8 whether the designated sectors are replaced by other sectors through the occurrence of defective sectors and gives the recording/reading device 4 an instruction to read suitable sectors. If a defective sector is detected when the writing operation is carried out according to the writing command, the controller 5 rewrites the information to a suitable replacement sector on the basis of the defect management information stored in the RAM 8, writes a new defect management information into a suitable sector, and simultaneously renews the data in the RAM 8 to the latest defect management information.
As described above, in the conventional autochanger it is necessary to read the defect management information from the optical disc every time the disc is installed into the recording/reading device 4. Therefore, it takes additional time corresponding to reading the defect management information to install the optical disk. The reason for necessitating the extra time to read out the defect management information will be described in the following.
FIG. 5 is an explanatory diagram of an example of a defect management method for a write-once type optical disc with a diameter of 130 mm. There are 20,000 tracks (track numbers are 1 to 20,000) in the disc and these tracks are used for a map area, a user area, or a replacement area. One track consists of 32 sectors and each sector has capacity of 512 bytes.
The user area is originally used by the user to write data into it and read data out of it. The replacement area is used to rewrite the data of the sector which is judged as a defective sector in verification reading performed immediately after the data have been written into the user area. A defective sector means a sector which has micro-defects, so that the data previously written can not be read out with a prescribed signal quality. The map area is used to write the defect management information which indicates the correspondence between the defective sectors in the user area and the replacement sectors in the replacement area.
The disc memory area can be divided into a plurality of bands consisting of at most 63 bands. Basically, each band has a map area, a replacement area and a user area. However, as described below, there exists a band which does not have a user area. The map area and the replacement area of each band are determined to have four tracks (128 sectors), respectively, but the number of tracks in each user area is not fixed. Also there is not determined the position of the first track of each area of each band. The number of tracks in each user area and the positions of the first tracks may be designated by the user or by the controller automatically. In order to memorize the above information, there are provided control tracks with the fixed track numbers (except for 1 to 20,000) in the disc.
FIG. 5 shows an example of a format of the band division and the track numbers of each area of the disc memory area. As shown in FIG. 5, the disc memory area is divided into the bands of #1 to #63. The bands of #1 to #62 have respective user areas and the band #63 does not have a user area. Each user area of the bands #1 to #61 consists of 314 tracks and the user area of the band #62 has 342 tracks so as to adjust the fractions.
When data are written in a certain sector in the user area #1 and the sector is judged to be defective in the verification reading carried out just after the writing, the same data are again written into the sector with the smallest address number among non-written sectors in the replacement area #1. This replacement sector is also subjected to verification reading. When it is a defective sector, the data are again written into the right next sector. In this manner, the same data is written into the sector in the replacement area #1 repeatedly until the sector without defect is found in the verification reading. As a result, one sector in the replacement area #1 is given to one defective sector in the user area #1. One pair of addresses of these two sectors constructs one defect management information and this information is written into the sector with the smallest address number among non-written sectors in the map area #1. This sector in the map area #1 is also subjected to the verification reading. When this sector is found to be a defective sector, the data are again written in the right next sector. In this way, the same data (defect management information) is written into the map area #1 until the sector without defect is found in the verification reading.
The same procedure is carried for the bands #2 to #62. The band #63 is provided for the case of overflowing in the map area or in the replacement area of the bands #1 to #62. Therefore, the band #63 does not have a user area.
Construction of the map area will be explained in the following. One sector (512 bytes) in the map area is divided into 128 fields, with each field having four bytes. Three bytes in one field show the defective sector address in the user area and the remaining one byte shows the specific sector of 128 sectors in the replacement area, by which specific sector the defective sector in the user area is replaced. Therefore, the one sector in the map area can contain at the maximum 128 of defect management information and it is identical with the number of 128 sectors in the replacement area provided in one band, so that only one sector in the map area can contain all the defect management information of the corresponding band. When the data in a certain band are overflown, the band #63 as the common overflowing area is used to store the overflown data.
As data-rewriting cannot be performed in the write-once type optical disc, when one defect management information is generated in a certain band, it is unavoidable to consume one sector in the map area, into which sector the defect management information is written. When the defective sector is found at the first time in a certain band, the data describing the defect management information are written into the first field (4 bytes) of the leading sector in the map area of that band. And when the next defective sector is found, the same data describing the first found defect management information are written into the first field of the second sector and the data describing the newly found defect management information are written into the second field of the second sector (the second field of 4 bytes) in the map area, since the data in the leading sector of the map area cannot be rewritten.
As is understood from the above explanation, the latest defect management information in a certain band exists in the last written sector in 128 sectors in the map area of that band. Needless to say, the latest defect management information is necessary when reading the data out of the user area. Moreover, when a new defective sector is found in writing the data; without the latest defect management information it is uncertain into which sector in the replacement area the data should be rewritten, and without the knowledge of the sector address in the map area storing the latest defect management information it is also uncertain into which sector in the map area the next defect management information should be written.
Therefore, the following operations are needed when a disc is loaded in the recording/reading device. Firstly, the control tracks in the disc are read to know the track numbers of the map area of each band. Secondly, the last-written sectors are found out for all map areas and the latest management information of each band is read out of that sector. Finally, those of latest management information are written into the RAM.
Next, the time necessary to read the defect management information will be explained. If the rotation speed of the disc is 1,800 r.p.m., the time for one rotation of the disc is about 33 msec. Presuming that a considerable portion of the map area of each band has been consumed, it is hard to expect shortening the time by employing the process of finding the last-written sector in the map area of a certain band and immediately after that accessing to the map area in the next band. Eventually, there is no other way than reading all the map area of the disc according to the revolution of the disc. The total number of the tracks in the map areas is 252, so that 8.4 seconds are consumed to read all tracks in the map areas.
As in the autochanger for optical discs the disc-loading time is one of the important performances, the operating speed of the disc-loading mechanism must be made higher. However, in accordance with the prior art, more than 8 seconds of time is necessary as described above to read the defect management information, resulting in the disadvantage of increasing the disc-loading time.